116W.Shuwenetal./EcologicalEngineering18(2001)115–120

causedsuchfailures?Thecommonanswerisun-certaintyaboutwhatisnecessarytocreateandrestoretheseimportantecosystems(Roberts,1993;MitschandWilson,1996;Young,1996;Zedler,1996;Malakoff,1998;Mitschetal.,1998).Peoplespendmuchtimestudyingthecomplexinteractionsamongcomponentsofanecosystem,butstillfeelagapintheunderstandingofsuchinteractions.Ifthegoalofwetlandmitigationistomimiccloselythefunctionsoftheoriginalsystem,asmanyecologistsbelieveitshouldbe,thejobisstilldif cult(Roberts,1993).

Inwetlandcreation,wedesignthecommunitystructureandsomeoftheenvironmentalcondi-tions,andanticipatethatthesystemwilldevelopintoanexpectedone.Itisaprocessofsecondarysuccession.Whyisitsodif culttoacheive nalsuccess?Maybeweshould ndtheanswerfromthewhole-ecosystemdevelopmentratherthantheinteractionsinasystem.Thestructureofthecommunity,especiallytheplantcommunity,canbeusedasoneindicatorofmaturityofanecosys-tem.Accordingtothepolyclimaxviewpoint,theclimaxcommunityisthelocalend-pointofasere,andisdeterminedbylocalenvironmentalcondi-tions.Sowhetherwecandesigntheenvironmen-talconditionstobejustthesameastheoriginalsystemisofgreatimportancetoensurethechar-acterofthecommunity,andthusadesiredma-tureecosystem.Indeedwemaydoalotofworktomimiccloselytheenvironmentalconditionsoftheoriginalsystem.Butitseemsverydif cultbecausetheenvironmentalconditionsarecom-prisedofsomanyfactors,suchasclimate,soilnutrients,moisture,slope,exposure, re,etc.Forexample,theprobleminacreatedcoastalwetlandinSanDiegoBaywasinadequatenitrogensup-pliesbecauseofsandysoils,whichpreventedcordgrassfromreachingitsfullheight;thusthecreatedwetlandfailedtoduplicatethefunctionsofanaturalone(Roberts,1993).InIllinois,7-year-oldcreateddeepwatermarsheslackedsimi-laritytonaturalmarshesbecauseofarti ciallydeepwaterandlackofsoilcarboninthecreatedwetlands(MitschandFlanagan,1997).

Moreover,designingthecommunitystructurecompletelyliketheoriginalonealsoseemsimpos-sible.Whittaker(1967)emphasizedinhisclimax-

patterntheorythattheconstituentpopulationsinaclimaxcommunityareindynamicbalancewiththewholepatternofenvironmentalfactorsinwhichitexists—‘climate,soil, re,bioticfactors,andwind’.Thecommunitystructureisacontin-uumofoverlappingsetsofspecieseachrespond-ingtosubtlydifferentenvironmentalcues,andthusnode nitepatternoftheoriginalsystemcanbeimitated.‘‘Whileitiseasyto gureoutwhichplantstobringin,wheretoputthem‘speci cally,atwhatelevation’isnotsoclear.Plantingthemafewinchestoohighortoolow,inrelationtothetidalregime,canspelldeathtoanewlyintroducedplantpopulation,’’saysZedler(Roberts,1993).

2.Designwetlandcreationaccordingtothe

principlesofecologicalsuccessionandleavemorespacetoself-design

Humansstillcannotdesignanecosystemasdelicatelyasnaturedoes.Whentoomanyhumanstandardsdonotseemtowork,weshouldleavemorespacetonatureandgiveherachancetoexpressherself(Mitsch,1993).Peoplehavebeguntopaymoreattentiontothenaturedesignthanhumandesign.Awhole-ecosystemexperimentconductedinTheOlentangyRiverWetlandRe-searchParkatTheOhioStateUniversityalsocon rmedthegreatself-designabilityofnature,andthetheoryofself-designforwetlandcreationandrestorationwasbroughtout(Mitschetal.,1998).However,howcanweletnaturedevelopinthewaywedesirefromthebeginning?Naturecannotconsiderallofourrequests,andsheworksindependentlyaccordingtoherrules,theprinci-plesofecologicalsuccession.Soifwewanttoworkwithnaturetowardthesamegoal,wemustobeytheprinciplesofecologicalsuccessionindesigningourprojects,andleavemorespaceforself-design.

Comparedwiththetraditionalway,moreef-fortsareputtochoosingapropersitetoensurethesystem’sdevelopingdirection.Atleastthreeaspectsshouldbeconsideredwhenchoosingasite:(1)abilitytochangeintotheintendedsystem,whenanalyzedfromsuccessionhistory;(2)the